板塊構造：地殼循環的演化

原文網址：https://www.nature.com/ngeo/journal/vaop/ncurrent/full/ngeo3015.html

板塊構造：地殼循環的演化

Valentina Magni

大陸地殼形成和循環的過程會隨著時間經過產生變化。數值模型顯示此過程從早期地球經由下部地殼剝離導致的大範圍循環，演變成今日藉著隱沒板塊斷裂形成的局部循環。

在整部地球歷史中，大陸地殼經歷著形成與消滅的過程。科學家對此關係的消長有相當激烈的辯論，但他們已經漸漸同意大部分的大陸地殼是在地球歷史早期形成，而且從那時開始就有部分會循環回地函當中。要瞭解過去是由什麼作用主導了地殼循環並非易事，因為我們對於早期地球板塊運動的類型和力學原理擁有的資訊十分稀少。Chowdhury 等人在《自然―地質科學》(Nature Geoscience )撰寫的論文中，利用數值模擬顯示地球在太古宙晚期和元古宙早期溫度較高的時候，地殼循環的特性為緻密的下部大陸地殼剝離而沉入地函之中，其規模和速度可能比今日高上許多。

隱沒帶是今日大陸地殼形成和消滅的地點。大陸地殼主要形成於隱沒帶上方的火山弧，此處產生玄武岩質岩漿的成分因為接下來的分化作用而偏往矽鋁質。另一方面，板塊碰撞時透過隱沒作用和陸地物質的侵蝕作用，可以讓相較之下較少量的大陸地殼循環回地函。隱沒作用也是板塊運動的驅動力之一。因此，過去數十年來自然有大量的研究焦點放在瞭解隱沒作用的力學原理，以及更全面的，瞭解板塊的運動方式如何演變。之前從太古宙陸地矽鋁質的部分中發現地球化學訊號呈現出火山弧岩漿的特徵，讓有些學者提出太古宙時大陸地殼也是經由隱沒作用形成的說法。但有人則提出跟現今同樣類型的板塊運動要到地球歷史稍晚才開始運作，因此早期的大陸地殼可能是由不同機制產生。舉例來說，地球早期厚重且緻密的鎂鐵質地殼陷入地函當中，接著發生地函熔融產生鎂鐵質岩漿，隨後分異作用使得較為矽鋁質的地殼得以形成。雖然對於形成早期大陸地殼的作用還莫衷一是，但我們知道跟現今矽含量較多的陸地相比，它們的成分平均而言有更多鎂鐵質。

最近陸地增長的模型顯示現今大陸地殼的60%至70%大約30億年前就已經出現了。然而，該年代的地殼中只有不到10%留存至今日，意謂早期大陸地殼有很大一部份已經消失回地函當中。

Chowdhury和他的同事提出一種機制可以同時解釋發生在太古宙晚期和元古宙早期的大規模地殼循環，以及較為矽鋁質的大陸地殼如何形成。他們運用數值模型來模擬在不同溫度條件下，也就是地球歷史不同時刻的大陸碰撞過程。他們的模擬結果顯示在幼年地球具有的高溫地函環境下，會比今日更容易發生「層脫」(delamination)作用。此作用的特點是密度較高的下部大陸地殼跟岩石圈地函會脫落並沉到地函內部，剩下密度較低的上部大陸地殼留在表面。如果此種作用發生的範圍相當廣大，那它就有機會將大量下部大陸地殼迅速循環回地函之中。不僅如此，層脫作用也能引發地函上湧而產生大範圍的矽鋁質岩漿活動，促使矽質陸塊形成。

圖1：板塊運動和碰撞帶大陸地殼循環的類型演變

a. Chowdhury和其同事利用數值模擬顯示年輕、高溫地球的特點為具有層脫類型的板塊運動。下部大陸地殼和岩石圈地函會剝離而將大量下部地殼循環回地函當中。b. 相較來說，在現今條件下更常發生的是隱沒板塊斷裂，但此作用只能將少量的上部和下部大陸地殼循環回去。

隨著地球地函的溫度逐漸下降，循環類型可能也跟著改變。今日某些碰撞帶仍會發生層脫作用，但跟現今隱沒作用發生之後的隱沒板塊斷裂相比少見許多。有趣的是，因為只有一小部分的大陸地殼會隨著海洋板塊脫落，因此隱沒板塊斷裂循環大陸地殼的能力跟層脫作用相比小了非常多。故模型顯示今日地球相比早期地球可以更容易地保住大陸地殼。

板塊碰撞的力學原理和地殼循環的方式隨時間改變造成了一些重要後果。舉例來說，有人認為早期陸地因為成分偏鎂鐵質且漂浮於強度較弱的地函之上，所以不會露出海平面。但由於層脫作用去除了岩石圈地函和地殼密度較高的鎂鐵質部分，加上有更多矽鋁質物質產生，使得形成的陸地比較輕而可以升至海面以上。露出海面的陸地會改變風化、侵蝕作用和微生物的居住環境，對某些重要的揮發性元素，比方說氧的循環過程造成劇烈變化。Chowdhury和其同事的模擬結果也確實顯示出經由層脫作用發生的地殼循環範圍相當廣泛，且陸地矽化過程達到的高峰也跟大約在24至21億年前發生的大氧化事件差不多同時。

由剝離過程主導的循環類型和隱沒板塊斷裂主導的循環類型有個有趣的差異：進入地函的物質種類，並連帶影響地函化學成分。層脫作用發生時鎂鐵質的上部地殼會留在地表；但隱沒板塊斷裂時則有一部份的上部地殼循環回地函。在今日的對流地函當中，此意謂著上部地殼的碎片可以進入上升的地函柱，跟地函一起熔融而在中洋脊形成富化成分的海洋新地殼。

Chowdhury 等人利用估計大陸地殼循環速率的變化，提出大約是在10億年前板塊運動的方式轉變成跟現今一樣。然而，此估計值會受到數值模型內建的假設以及跟地函溫度演變相關的不確定性影響，因此要明確指出板塊運動開始變成我們今日所知樣貌的時間點並不容易。更重要的是，我們必須記住板塊增生的速率不只會隨著時間變化，在個別碰撞帶之間的空間差異也有巨大的差別。對於特定時間點只利用單獨一個循環速率的平均估計值來代表可能無法完整呈現碰撞帶的複雜性和多變性。不過這項新的預估結果還是提供了一個重要的切入點，讓我們能探討全球尺度下地殼循環過程的生涯變化。

Plate tectonics: Crustal recycling evolution

The processes that form and recycle continental crust have changed through time. Numerical models reveal an evolution from extensive recycling on early Earth as the lower crust peeled away, to limited recycling via slab break-off today.

Continental crust has been created and destroyed throughout Earth's history. Estimates of the balance between this loss and gain are vigorously debated, but there is a mounting consensus that a large portion of continental crust formed early in Earth's history and has since been partly recycled back into the mantle^1, 2. Determining which processes were responsible for crustal recycling in the past is a difficult task because we have very little information on the dynamics and style of plate tectonics on early Earth. Writing in Nature Geoscience Chowdhury et al.³ use numerical simulations to show that when Earth was hotter — during the late Archaean and early Proterozoic — crustal recycling was probably much more extensive and rapid than today, and characterized by the peeling-off of dense lower continental crust into the mantle.

Today, subduction zones are sites of continental crust formation and destruction^2, 4. Continental crust is mainly created in volcanic arcs above subduction zones, where basaltic magma generates and subsequently differentiates to more felsic compositions. Relatively small amounts of continental crust are recycled back into the mantle as the tectonic plates collide, through subduction and erosion of continental material. Subduction is also a driver for plate tectonics. It is therefore natural that in past decades there has been considerable focus on understanding how subduction dynamics and, more generally, plate tectonic style has evolved. Observations of the geochemical signature of volcanic-arc magmas found in the felsic parts of Archaean continents has led to the idea that continental crust formed via subduction during the Archaean, too^5, 6. Yet some suggest that modern-style plate tectonics did not begin on Earth until later⁷ and that early continental crust might have formed by different mechanisms. For example, foundering of thick, dense mafic crust into the mantle followed by mantle melting and consequent fractionation of mafic magmas to create more felsic crust⁸. Although there is no agreement on the processes leading to its formation, we know that early continental crust was on average more mafic compared to the present-day continents that are more silica-rich⁹.

Recent continental growth models suggest that 60 to 70% of the present-day volume of continental crust was already present about three billion years ago¹. However, only less than 10% of crust of that age is preserved today, implying that a large portion of the early continental crust has been lost back into the mantle.

Chowdhury and colleagues³ propose a mechanism that can explain both the extensive crustal recycling and the formation of a more felsic continental crust during the late Archaean and early Proterozoic. They use numerical modelling to simulate continental collision at different mantle temperatures; that is, at different times in Earth's history. Their simulations show that under the much hotter mantle conditions that characterized the younger Earth, delamination would be easier to achieve than today. This process would have been characterized by the peeling-off and sinking into the mantle of the dense lower continental crust, and sub-lithospheric mantle from the less dense upper continental crust, which remained at the surface (Fig. 1a). If widespread, then this process would have had the potential to recycle large amounts of lower continental crust rapidly. Moreover, this delamination could have triggered mantle upwelling that produced extensive felsic magmatism, leading to the formation of silicic continents.

Figure 1: The evolving style of plate tectonics and continental crustal recycling in collision zones.

a, Chowdhury and colleagues³ use numerical simulations to show that the younger, hotter Earth could have been characterized by a delamination style of plate tectonics. The lower continental crust and sub-lithospheric mantle could have peeled away, recycling large volumes of lower continental crust into the mantle. b, In contrast, under present-day conditions slab break-off is common, but this process only recycles small fragments of upper and lower continental crust.

As Earth's mantle temperature gradually decreased, the recycling style probably changed. Today, delamination still occurs in some collision zones, but it is much less common than modern-style subduction followed by slab break-off. Interestingly, the recycling potential of slab break-off is considerably smaller than for delamination because only a fragment of the continental crust detaches with the oceanic plate (Fig. 1b). Therefore, the model results show that it is much easier to preserve continental crust today than it was on early Earth.

The changing style of collisional dynamics and crustal recycling through time has some crucial consequences. For instance, it is thought that early continents, being more mafic and supported by a weaker mantle, would not have been above sea level. The removal of the sub-lithospheric mantle and the denser mafic part of the crust due to delamination, in addition to the production of more felsic material, would create lighter continents that could rise above sea level. The sub-aerial emergence of the continents would lead to changes in weathering, erosion and microbial habitation that would drastically change the cycles of some key volatile elements such as oxygen. Indeed, the model results from Chowdhury and colleagues suggest that recycling via delamination would have been widespread and the silicification of the continents would have reached a peak at a time roughly coincident with the Great Oxidation Event about 2.4 to 2.1 billion years ago¹⁰.

An interesting difference between a predominantly peeling-off style of recycling compared to a slab break-off style of recycling is the type of material that goes into the mantle, and thus the influence on mantle chemical composition. During delamination, the felsic upper crust remains at the surface, whereas with slab break-off a portion of the upper crust is recycled. In the present-day convecting mantle, this means that fragments of upper crust could be entrained into upwelling mantle plumes and melt together with the mantle to produce new oceanic crust at mid-ocean ridges with an enriched component.

Chowdhury et al.³ use estimates of the changes in continental crustal recycling flux to suggest that the transition to modern-style plate tectonics happened about one billion years ago. However, these estimates are affected by the intrinsic assumptions in numerical models and uncertainties related to the evolution of mantle temperature, so it is difficult to pinpoint when plate tectonics started in the form we know it today. More importantly, we have to keep in mind that the rate of crustal growth not only varies through time but also varies considerably in space, among individual collision zones. Using one average estimate of recycling flux as characteristic of a particular time might not fully represent the complexity and variability of collision zones. Nevertheless, these new estimates are an important starting point to investigate the fate of recycled crust at a global scale.

原始論文：Chowdhury, P., Gerya, T & Chakraborty, S. Emergence of silicic continents as the lower crust peels off on a hot plate-tectonic Earth. Nature Geoscience, 2017; DOI: 10.1038/ngeo3010